Different read counts

[oomvel]

Hello all,

I am Velmurugan from India working on the role of miRNAs during environmental chemicals accumulation in mammals. In this regard, we have done illumina small RNA sequencing from control and environmental chemical treated animal. I received different read counts in the two samples. Control - 26 Million reads and treated one - 39 million reads.

My question is whether these reads with huge variation in read count (26 M vs 39 M) is reliable for studying the differential expression analysis?

Thank you,

Velmurugan

[westerman]

You'll have to use statistics (or simple arithmetic if that is acceptable) to correct for the different numbers however the difference will be able to produce "reliable" results. I am more concerned about the lack of biological replicates. With only one control and only one treated animal it is hard to make definitive conclusions. In other words even if both of your samples had the exact same number of reads the reliability of your conclusions will be suspect.

[Geneus]

You'll have to use statistics (or simple arithmetic if that is acceptable) to correct for the different numbers however the difference will be able to produce "reliable" results. I am more concerned about the lack of biological replicates. With only one control and only one treated animal it is hard to make definitive conclusions. In other words even if both of your samples had the exact same number of reads the reliability of your conclusions will be suspect.

What's wrong with a N=1?

[oomvel]

You'll have to use statistics (or simple arithmetic if that is acceptable) to correct for the different numbers however the difference will be able to produce "reliable" results. I am more concerned about the lack of biological replicates. With only one control and only one treated animal it is hard to make definitive conclusions. In other words even if both of your samples had the exact same number of reads the reliability of your conclusions will be suspect.

Dear Westerman,

Thank you for your explanation. From my knowledge, most of the NGS studies in animal experiments were done in one sample (from each group) and after analysis, the results are validated by qPCR in 6 animal tissues.

Hope my strategy is correct.